Evrim Guler

High‐dimensional financial machine learning (ML) pipelines are computational workloads as much as predictive models: their practical value depends on runtime, memory footprint, scalability, and the ability to retrain under resource constraints. This paper treats empirical asset pricing as a demanding real‐world workload and proposes Cost‐Aware Adaptive Optimization , using XGBoost (extreme gradient boosting), as a systems‐aware framework to improve the computational efficiency of gradient‐boosted tree pipelines while preserving application‐level predictive utility. The framework uses computational feedback; optimization is driven by training time, prediction time, memory footprint, feature dimensionality, temporal training‐window depth, model complexity, and marginal utility per unit of computation. We first profile 10 ML algorithms in a controlled high‐performance computing setting and decompose execution cost into training and prediction components. XGBoost is then selected as the optimization target because it offers a strong cost–utility baseline on the asset‐pricing workload. The proposed framework adaptively controls feature‐dimensionality and temporal‐depth choices, evaluates memory‐ and execution‐aware trade‐offs, and analyzes scalability through runtime growth, parallel speed‐up, and parallel efficiency. Application‐level utility measures, including out‐of‐sample R ² , mean squared error (MSE), and the Sharpe ratio of decile‐sorted long–short portfolios, are retained as application‐level validation metrics rather than as the sole objective. The results show that meaningful computational savings can be obtained by treating dimensionality, temporal depth, and model complexity as controllable workload variables. The study therefore contributes a computational feedback‐driven optimization framework for scalable financial ML execution, framing workload adaptation and configuration selection as runtime‐, memory‐, and utility‐aware system problems.

Peer-to-peer (P2P) energy trading, Smart Grids (SG), and electric vehicle energy management are integral components of the Internet of Energy (IoE) field. The integration of Software-Defined Networks (SDNs) and Blockchain (BC) technologies into the IoE domain offers potential benefits that have only been studied in the literature in a few works. In this paper, we investigate the state-of-art solutions that leverage both SDNs and blockchain within the realm of the IoE. We categorize these solutions based on the method of integrating SDN and BC into two categories. The first category is the blockchain for SDN, where blockchain enhances the SDN directly. The second category is blockchain and SDN, where both technologies are used to enhance the proposed solutions. We identify three distinct blockchain applications based on their usage: decentralizing the SDN control plane, serving as a decentralized platform, and improving security measures. Similarly, we observe that SDN serves as a performance enhancer, a substitute for traditional networking, and solely as a control and management framework. It is posited that integrating SDNs and blockchain into IoE leads to performance enhancements, improves security, enables decentralized operations, and eliminates single points of failure in the SDN control plane. Additionally, some unaddressed issues, such as energy efficiency, smart contract management, and scalability, are discussed as potential future directions.

In the 5G and beyond 5G networks, achieving security-aware data transmission needs to convert clients ’ requests into a service function chain (SFC), each service function (SF) providing a certain security guarantee. With diverse configuration techniques, an SF may own multiple versions, each version providing various security guarantees with diverse costs. It should be notice that, as the recent software failures have caused severe financial loss, great attentions from both academia and industry have been put onto the SFC reliability. In the literature, existing works have solely investigated the following two fields: 1) how to deploy a security-aware SFC, and 2) how to protect a traditional SFC. Simply applying these techniques to dealing with the problem of security-aware SFC protection might not be efficient as the backup and primary SFCs may not be identical for security-aware SFCs. Therefore, how to jointly take these fields into account is challenging and remains open. To tackle the above problem, this paper studies how to construct and embed a security-aware SFC with asymmetric dedicated protection. We mathematically define this problem and name it security-aware service function chaining, embedding, and protection with multi-versioned SFs (SFCEP-MF) with the objective of cost optimization. Next, to optimize the SFCEP-MF problem, we construct an efficient algorithm, called augmenting-path with primary-first disjoint SFP identifier (APPF-DSI). Extensive simulation results show that the APPF-DSI algorithm outperforms the benchmark approaches that are directly extended from the state-of-the-art.

The immense potential of optical networks provides a highly favorable option for addressing rapidly increasing bandwidth needs. Elastic Optical Networks (EONs) exhibit considerable potential in enhancing availability, failure resilience, load balancing, and efficient resource allocations, rendering them a viable option for forthcoming high-speed networks. The integration of Software Defined Networking (SDN) architecture with EONs, which separates the data and control planes, results in a dynamic and cooperative combination known as Software Defined Optical Networking (SDON). However, the effective allocation of spectrum poses operational difficulties in this context. The proposed study introduces a framework called Cross-ISP (Internet Service Provider) Traffic Engineering enhanced by Particle Swarm Optimization (CITE-PSO) that is designed to facilitate Quality of Service (QoS)-focused cross-ISP spectrum assignment in SDONs through the integration of blockchain technology. The proposed framework aims to eliminate the need for centralized mediators while ensuring effective coordination of inter-ISP traffic based on QoS considerations. The research presents a novel application of blockchain technology that increases network efficiency and minimizes QoS signaling overhead during inter-ISP routing in SDONs. The proposed framework is evaluated under Hop-by-HopWavelength Switching (HWS) and Border-Node-Only Wavelength Switching (BWS) to assess its performance. The simulation results demonstrate that the CITE-PSO framework is proficient in managing the inter-ISP routing with QoS capabilities in the SDON architecture. This proficiency is measured by the various metrics across HWS and BWS scenarios.

Bu çalışma, acil uzaktan eğitim sırasında ve sonrasında değerlendirmeye ilişkin akademisyenlerin algılarını ve uygulamalarını araştırmayı ve karşılaştırmayı amaçlamaktadır. Araştırma kapsamında, iki farklı zaman dilimindeki akademisyen deneyimlerini, zorluklarını ve ihtiyaçlarını anlamaya odaklanılmıştır. İlk aşamada, pandemi nedeniyle acil uzaktan eğitimin uygulandığı Temmuz ve Ağustos 2020'de çeşitli disiplinlerden 54 öğretim üyesiyle 19 yarı yapılandırılmış odak grup görüşmesi gerçekleştirilmiştir. Sonraki aşamada, pandemi koşullarının ortadan kalkmasının ardından ayrı bir grup öğretim üyesiyle (4 kişi) görüşülmüştür. Görüşmelerden elde veriler MaxQDA yardımıyla nitel tematik analiz yaklaşımı kullanılarak analiz edilmiştir. Çalışmanın bulguları, acil durum uzaktan öğretimi sırasında değerlendirme uygulamalarının akademisyenleri mesleki gelişimi, öğrenci motivasyonu ve Öğrenme Yönetim Sistemlerinin işlevselliğinden güçlü bir şekilde etkilendiğini göstermektedir. Ayrıca, çalışma, akademisyenlerin değerlendirme alışkanlıklarının ve uygulamalarının acil uzaktan öğretim sürecinden sonra değiştiğini ortaya koymaktadır. Bu bulgulara dayanarak, yükseköğretim kurumlarına mesleki gelişim fırsatlarını artırmaları ve değişen değerlendirme uygulamalarına uyum sağlamada akademisyenleri desteklemeleri için önerilerde bulunulmuştur.

The truly innovative network virtualization technology allows for multi-tenancy, enabling different Virtual Network (VN) requests to share the same physical network. This is achievable because network services are no longer bound to the architecture of the lessor hardware. Virtual Network Embedding (VNE) is a multi-dimensional NP-Hard problem that maps VN entities such as virtual nodes and virtual links onto a shared Substrate Network (SN) while assuring the requested network resources (e.g., bandwidth, computing power, etc.). This research explores how to efficiently map VNs with one-to-many (multicast) interactions, in the form of Virtual Multicast Trees (VMTs), onto an SN in contrast to the VNE problem where one-to-one (unicast) communication is at focus. Thus, we propose a Virtual Multicast Tree Embedding (VMTE) enhanced by Particle Swarm Optimization framework, VMTE-PSO, to put VMTs on a shared SN. The VMTE-PSO aims to minimize the amount of network resource consumption (i.e., bandwidth) in the SN while simultaneously satisfying the computing demand of virtual nodes and minimizing the redundant substrate link usage. Extensive simulations reveal our algorithm outperforms the dynamic node ranking and traditional greedy-based VMTE approaches with respect to bandwidth consumption and redundant multicast transmission on NSFNET and USNET network topologies.

The state of computer network technologies has continually advanced at a rapid pace. Software Defined Networking (SDN) and Blockchain (BC) have emerged as complementary technologies providing support that facilitates greater security and greater network performance for many domains of application, including the Internet of Things (IoT) ecosystem, ideally resulting in an improvement in our collective quality of life. The proliferation of IoT devices, driven by a wide variety of use cases and its ubiquitous availability, combined with the emergence of SDN and BC, presents rich opportunities for various emerging research efforts. This paper presents a comprehensive survey of the studies in which BC and SDN have been integrated into the IoT ecosystem, referred to hereafter as BC-enabled Software- Defined IoT (BC-SDIoT). First, we discuss the motivations and drivers for integrating BC-enabled SDN and BC-SDIoT, as well as the benefits and drawbacks. Second, we categorize the relevant studies according to six key implementation objectives and ideas that combine BC, SDN, and IoT technologies to create smart, secure, and effective frameworks: Security, computing paradigms (edge and fog computing), trust management, access control & authentication, privacy, and networking. In the corresponding sections, we present the categories (i.e., problem domains) of the aforementioned novel taxonomy and discuss related studies (i.e., solutions) in depth. Finally, we outline potential major challenges, open issues, and future prospects that require further research attention and intensive endeavors for complete and ground-breaking frameworks to broaden newer research domains in BC-SDIoT. This survey paper may be a fruitful primer for a reader investigating the exploitation of BC in SDN and IoT ecosystems.